A Kewa Dictionary

44

3.1. WOUVS-^

3. 1 .0, Introduction

In this section both word classes and word patterns are outlined.
If both the members and morphological combinations of a given class are
restricted, they are described In the same section.

3. 1. K Worti trasses

Two major groups of words may be conveniently defined In Kewa: Verbs
and Non-Verbs. The criteria employed to classify verbs are primarily
Inflectional coupled with morphophonemlc variations of the verb bases
and certain suffixes. This results In a clear division between a class
of verbs on the one hand and all non-verbs on the other. The sub-
division of non-verbs is less clear-cut, but several additional word
classes can be substantiated: Nouns, Adjectives, Adverbs, Deictics ,
and possibly a class called simply Particles. The latter Includes
several clitics which are not clearly words because they do not occur
as free forms. On the other hand, clitics cannot be considered as
affixes because at least some of them can occur, in combination with
each other, as words. The classification which follows is not exhaus-
tive, nor is this to be expected realistically for any first description

2
of a language. However, by giving definitions of word classes the

category symbols employed for them throughout the grammar should be
clearer to the reader.

3.1.1.1. Ve^bA (v)

Verbs are those stems which may occur with (1) tense suffixes; (2)
the pre-clltlc na- '(negative) ' ; (3) command suffixes. Thus the defini-
tion of a verb may be on inflectional grounds alone. By incorporating
syntactic. I.e. additional distribution criteria, verbs also: (4) have
tense suffixes which are potentially in cross-reference with free pronoun
subjects; (5) occur with coordinating suffixes.

This section first appesired as Chapter 3 in A GAamtUi o^ Kewa, Wew Gooiea C-l6, 1971,
and is reprinted here by permission, with minor alterations necessitated by book
publication.

2
Matthews (1966:156) notes that exhaust iveness is only reqmred when word classes
are conceived as a taxonomlc system. He also makes two further relevant observations
about the notion of word classes: (l) definitions can naturally be heterogeneous,
i.e. employ 'notional' as well as 'fonnal* evidence, 'morphological' as well as
'syntactic' criteria; (2) at least some of the definitions should refer to universal
properties of grammars {ihid* t pp. 156-9).

^5

Verb stems are inherently Active (va) or Stative (vs). Stative
stems are signalled phonologically by bases ending only in /aa/ and
morphologically by occurring only with affixes of Set II (outlined in
3.2.2.1,). Stative verbs may be subclassified into non-derived and
derived stems, the latter arising from underlying active bases. Some
examples of non-derived stative stems are:

(1) nfmfnaa 'to understand'

(2) rumaa 'to portion out'

(3) kiritaa 'to gather together'

(4) mi nasaa 'to lift up'

Active stems have bases with phonemic shapes which end in /a/, or
in two cases /u/; they occur with affixes of either Set I or Set II.
Active bases which occur with Set II are derived stative stems.

3.1.1.2. Hoan^ (n)

The definition of a noun stem is essentially a negative one which in
fact applies to any non-verb , non-derived stem which cannot occur with
terminal or non-terminal suffixes (outlined in 3.2.2.2.). A positive
definition must employ functional criteria: those stems which potent-
ially expound the Head of a nominal phrase are nouns. When nouns ex-
pound functions such as SubJect-as-Agent or Object-as-Location of
certain clause types they are marked by special clitics. However, so
are other word classes, so there appears to be little or no morpholog-
ical evidence to separate nouns from other word classes, excluding
verbs.

When a noun expounds the Head-as-Agent of a descriptive noun phrase
(np, ) it may be modified and the full np is marked by the clitic -me

('V -mf). This particular clitic marks the function of Subject-as-Agent
in certain clause types. For example, the following are np_ marked

by -me (throughout, a hyphen indicates morpheme boundaries):

(5) adaa ii-mi 'the big man'

(6) adaa aa laapo-me 'the two big men'

Such clitics as -me mark clause-level functions for various gramma-
tical categories, not simply nouns. For this reason they are discussed

2
in Chapter 4 on clause patterns.

In other Highland languages possessive suffixes often divide noun stems into such
categories as kinship and kin or body parts and functions on the one hand, and
animate-inaniinate nouns on the other. Cf. for example, D. Bee (1973) on Usarufa for
the former, or P. Healey (1965:6) for the latter.

2

The Chapters cited in this section refer to K.J. Franklin, A GnamnoA. o^ Kewtt, Wcw
GtUma, PL C-I6, vhich should be consulted for further detail.

46

It is possible to sub-divide nouns semantlcally into those which are

general (ng) and those which are temporal (nt). Any further sub-

categorisatlon of ng is also on a semantic basis: for example, nouns

which are animate (ng ) versus those which are inanimate (ng. ). As

^an mam

will be shown later, an animate-inanimate dichotomy is reflected in the
semantic interpretation of functions such as SubJect-as-Agent on the
one hand, and SubJect-as-Instrument on the other. In addition, inter-
rogative pronouns (Cf, 3.1.1.5.1.) have animate-inanimate forms and
consequently support the division of ng into these two semantic cat-
egories.

Stems which specify time are a sub-class of nouns and include forms
such as: ibfa 'now*; rfbaa 'night'; orope 'later'; and oge oro
'Saturday ' .

Reference to specific points of time in the past or future is often
by a combination of nt and demonstratives (3.1.1.5.2.). Note, for
example, the bottom three rows in Chart 2.

Chart 2: Time Orientation

abfa 'now'

'yeate

rday ' a b a 1 a

ikirii '

tomorrow '

P

H

R

numane

'one day '

' J '
rudane

E

I

mo numane

'two daye'

mo rudane

N

apo numane

'three days'

apo rudane

C

R

E

However, any nt can also occur with substantive (= non-verb) clitics,
e.g. :

(7) ek^reia-mc. . . 'with (the coming of) tomorrow.,,'

(8) ekeraa-nane 'in the future*

(9) ekeraa-nu 'later on' (lit. , tomorrow -many) '

There seems to be no basis for a sub-classification of ng into common
versus proper.

In E. Keva names given to females are optionally suffixed "by -nyu or -me: Rumu-nyu
'(the woman) Rimtnyu'j Warua-me '(the woman) Waruame'. In addition, any name may be
suffixed to shov the parentage: Rumunyu-raa 'the father of Rummy u* ^ waruame-gi 'the
mother of Vareame'^ where the suffixes -raa and -gi are contractions of the kinship
terms of reference aaraa 'father* and agi 'mother', (Cf, K. Franklin 196Ta:T8).
Although both types of suf fixation occur In W, Kewa, the system appears to be less
developed. In W. Kewa female names are frequently formed by compounds employing
nogo *girl*j e.g. Kadfpf-nogo, Amala-nogo, Ipare-nogo.

^7

3.1.1.3. AdjZctJ^Ve,^ (aj)

Steins which may expound the Modification function of an np or often
the Complement function in a clause are non-derived adjectives. Such
stems may be sub-categorised on the basis of their potential position
and function in a noun phrase. The following sub-categories of adject-
ives will be described in greater detail later in Chapter 5 on Phrases:

(a) colours (aj , ) are forms such as kobere 'dark 'j

kaane 'red', and kaga rekene 'green',

(b) size (aj^_) are forms such as adaa 'big', oge

sz

'little * , and rtjdu 'short'.

(c) number (aj ) are forms such as li^po 'two', ekataa

nm "^

2
'little finger', and paapu 'one oyote',

(d) quality (aj ,) are such forms as ^p^ 'good', wae

'bad', and purt 'atrong'.

Any verb also become a derived adjectival form upon filling the
modification slot of a np and by being marked by the suffixes -ne

(egocentric benefaction"^) or -ae (altercentrlc benefaction) (see
3.2.6.1.).

An aj may also be marked with the clitic -pu which specifies the
nm "^ *^

category of quantifier, usually with the meaning of 'sequence' or

'progression' , In addition, words of other classes or even sub-classes

of aj may become a derived aj :

nm

(10) nimu adaa-pu fmi {they all, big-quan, they have «

'they have plenty') [aj + -pu]

(11) ekeraa yaipi-pu fpa yotea (tomorrow, day~quan, water,

it will pull = 'tomorrow during the day the water
will recede') [nt + -pu]

(12) aki-pu rapoyaa (what-quan, quantity? = 'how many are

there?') [interrogative + -pu]

It is interesting to note that P. Healey (l965a:15-8) describes four sub-clsisses of
qualifier [=adjectives] which are very similar: colour, size, quality and quantifiers.
The latter is subdivided into genereO. and kin types.

2

See K. and J. Franklin (l962a) and also Franklin (l968b) for a description of how
body parts are used 6ls a system of counting in Kewa. (See also Section 11.3.).

3

Categories of benefaction are described in the section <^n verb affixation. Certain
morphophonemic rules outlined there also apply when adjectives are derived from verbs.

48

3.1.1.4. kdve.Kb6 (adv)

Stems which expound the Adjunct tagmeme (Chapter 4) of a clause
generally are adverbs. Adverbs do not occur with clitics which
typically mark functions expounded by nouns or adjectives. Such stems
may be derived or non-derived. The latter are a small closed set of
forms including: ai papulu 'quickly ' ; gupa 'likewise '; kodc 'almoet' ;
mada 'enough*; waru 'really'; a 1 oma 'quickly' (only with verbs of
motion); pawa 'slowly'; wala 'again'; and aba 'before*. Derived
adverbs (actually adverbials) are a combination of any stem which is
not already an adverb plus the clitic -rupa ' iadverbialiser) ' or any
construction plus this clitic. The latter represent embedded adverbial
clauses (outlined in Chapter 4). Some examples of derived adverbs are:

(13) nf aa-rupa pfralua (Jj man^like, I will sit =

'I will sit like a man') [ng + -rupa]

(14) nf etaa war'-ni-rupa lagfaa (Z, food, prep are -h^-yz-

like, he told = 'He told me how food is prepared')
[Adjectival Clause + -rupa]

(15) nf mogopara* rupa meawa (J, that over there~like,

I got it = *I got (the one) similar to that (thing)
over there') [Deictic Phrase + -rupa]

(16) nipu kfrJ pa-rupa ta {he, laugh, just-like, he is =

'He is just laughing {without reason)) [Particle +
- rupa]

3.1.1.5. VUcUc6 (del)

Deictics are either pronouns or demonstratives and constitute a
closed class, i.e. they can be enumerated. They can also only be de-
fined semantically by reference to the speech event and to the parti-
cipants of the speech event,

3,1.1.5.1, Pronouns (pro)

There are two sets of pi'onouns: Personal and Interrogative. Per-
sonal pronouns are a small closed set which may represent actors or
objects. The complete set of personal pronouns is given in Chart 3-

See Jakobson (l95T), especially his discussion of shifters and their semantic
components .

49

Chart 3: Personal Pronouns

In the third person, singular or dual is ambiguous, but may be
explicitly stated by the addition of laapo 'two':

(17) nlpu laapo ada-pe {he, two, they two aaw =

'They two eaw it ')

However, except when free pronouns function as objects, the use of
laapo 'two' , Is omitted because pronouns functioning as subject are in
cross-reference with verbs. In the following exsimples the subject of
the verb is marked as dual or singular by two separate suffixes (which
also indicate number and tense):

(17a) nlpu ada-pe {he, they two aaw = 'They two eaw it')
(19b) nipu ada-a {he, he saw - 'He saw it*)

If suffixes ambiguously signal the person of the actor, this may be
clarified by the use of free pronouns;

(18) nimi ada-me 'You all saw it'
(18a) n i mu ada-me 'They eaw it'

where -me ambiguously marks 2nd or 3rd plural Past.

Some examples of clitics occurring with pronouns are: nipij-ni {he--
pos = 'his'); nimu-para {you all~loc = 'to them'); ne-si {you-dLm -
'little you'); saa-me {we two-AGbi ~ 'we two').

Interrogative substitutes are built on two forms: (1) aapl '{an-
imate)'^ and (2) ake '{inanimate)' meaning broadly 'who?' and 'what?'
respectively. Note the appropriate nouns which are used in answer to
questions containing the interrogative forms :

(19) [aapf]-mf ta-a '[Who^ hit it?'
(19a) [ia]-me ta-a 'iThe man'\ hit it'

(20) [ake3-me ta-a *iwhatli hit it?'
(20a) [repena]-me ta-a '\^The tree^ hit it'

By expanding these two forms with general clitics, or by the use of
a separate interrogative clitic, the full range of interrogatlves such
as 'which', 'how many', 'why', 'how' and so on may be constructed.
These will be outlined under word patterns.

50

3.1.1.5.2. Demonstratives (dem)

Demonstratives, similarly to pronouns, are a small closed set of
substitutes, but which point out actors or objects. The complete set
of simple stems Is given in Chart 4 and other dem compounds are built
upon them.

Chart 4: Demonstratives

so

'ug*

go

'apeaifio' (seen)

mo

'distant'

o

'neutral'

no

' down '

apo

'general ' (unseen)

3.1.1.6. PaKtlclt^ (par)

Particles are non-derived or derived forms which function mainly at
the clause and sentence-level as Subordinators or Connectors. Subord-
Inators are free forms such as rabu 'when' in the following examples:

(21) epalla rabu epe ta {he will come^ when, good it is =

'When he aomee it will be good')

Connectors are of two shapes, either as clitics or derived from these
forms plus the affirmative verb 'to be' ya:

-pare 'but, however' yapare
-pulu 'because ' yapulu

-pae 'perhaps ' yapae

The forms in the first column are used if the relationship signalled is
between two verbal clauses or between a verbal clause and some other
construction; those in the second column are used if the relationship
is between a complementive clause and a verbal clause. Note the follow-
ing contrast:

(22) era lae-pare napalua {true, you said, but, I will

not go - 'You spoke true but I will not go')
(22a) era yapare napalua 'True, however I will not go'

3.2. WORV PATTERNS

As already indicated, words consist of simple or, in some cases, der-
ived stems. These stems occur with various patterns of affixation or
clitic orders. Those patterns which are contrastive (in the tagmemic
sense where usually two structural differences are required - see
Chapter 1) are separate syntagmemes. The patterns described In this
section begin with the more complex ones of verbs and proceed to the
simpler patterns.

51

3.2.1. Verb Bases

While describing these. It will also be necessary to outline the
morphophonemic rules which apply to both verb bases and certain affixes.
Simple stative verb bases are unaffected by morphophonemic rules. Der-
ived stative stems are affected by one general morphophonemic rule:

vstMP-Rl XV -^ Xaa / ABase Set II,

where V represents the final vowel of an active verb base which becomes
/aa/ in the presence of affixes of Set II (3.2.1,1.). XV refers to the
pattern of the active verb base; it always has one of the following
underlying morphophonemic patterns:

1. Pattern L - bases ending with or consisting of the
shape la, represented morphophonemically as XL,

2. Pattern E - generally bases ending with the vowel
sequence la or ea, represented morphophonemically
as XE.

3. Pattern A - generally bases ending with the vowel a,
represented morphophonemically as XA.

k. Pattern H - any base with an underlying shape of

XVi(C)V2> where V]^ = u or I , and V2 » a. Thus bases
ending in ia (Pattern E) automatically belong to
Pattern H.

Pattern H provides for morphophonemic rules of vowel harmony which
in every case override rules normally effective in Patterns L, E or A.
Therefore, such rules are ordered and always apply last after any other
appropriate rule.

Orthographlcally, patterns are marked as L, E, A or H only when
phonemic and morphophonemic correspondences are not isomorphic. Some
examples of each pattern are:

L: yola 'to pull'; la 'to talk'; pogolaa Ho jump'.

E: rla 'to aarry {on the ehoulder)' ; sa(E) 'to knit'; rakepea
'to huek'; 6geya(E) 'to beg'.

A: ada 'to look'; pu(A) 'to go'; mca(A) 'to fetoh'.

H: puna 'to shepherd* (basic pattern A); rabula 'to mend'
(basic pattern L); ta 'to hit* (basic pattern E).

■^In some cases if an alveopalatal or dental consonant occurs a high vowel preceding
it may be absent in present day Kewa. However, forms such as *ita 'to hit' can be
reconstructed.

52

MP rules operate upon such verb bases when they combine with oblig-
atory suffixes. The verb base plus the obligatory Terminal or Non-
Terminal suffix (see following section) constitute the obligatory verb
nucleus. Before discussing MP rules it is necessary to introduce the
obligatory suffixes which serve as conditioning environments to the
verb bases.

3.2.2. Obligatory Suffixes

Obligatory verb suffixes are either Terminal or Non-Terminal and in
each case belong to one of two sets: (1) Set I, which occurs only with
active verb bases, marks egocentric benefaction; (2) Set II, which occurs
either with stative or derived verb bases, marks altercentric bene-
faction. It is Important to note that the set marks either of the

categories of benefaction, although the individual affixes mark some

1
other grammatical category as well. Non-obligatory verb affixes co-
occur with Set I of Set II terminal or non-terminal suffixes, but do
not occur alone.

3.2.2.1. TtKmlnal Su^^-cxe^

Terminal suffixes mark either Tense or Imperative categories, which
are mutually exclusive with each other. Tenses are Present (Pr), Past
(Pa), Remote Past (RP), Future (Fu), and Perfect (Pf ) , Set II (alter-
centric) suffixes are outlined in Chart 5. Other abbreviations in this
section (KVM and KCM) refer to articles on verb morphology and clause
markers (Franklin 196^, 1965).

In KVM the eJ-trocentric Set is called Non-Personal Benef active. KCM introduces the
terms altercentric and egocentric but applies them only to tense. In this study
altercentric and egocentric apply to either terminal or non-terminal suffixes,
Typologically, the specification of egocentric or altercentric benefaction in verbal
categories is a distinctive characteristic of Kewa in particuleur, and the West-
Central Family in general. Wurm (1962:117) cites what he calls the use of "applicative
verb forms", i.e. "action for the sake of benefit of, a person other than the one
speaking, spoken to, or spoken about", as a typical feature of the West-Central Family:
In Telefol, of the Ok Family, P. Healey (I965c:6ff) describes a general dichotomy
between benefactive and non-benefactive stems. On the other hand, languages of the
Eastern Family have a benefactive morpheme which m\ist be preceded by an indirect
object meirker. (For Awa, see R. Loving and McKaughan 196U:19; for Gadsup, C. Frantz
and McKau^an 196U:d6; for Usarufa, D. Bee 196^:46.) In Benabena, a language of the
East Central Family, benefactive verbs take indirect object prefixes but are part of
a compound unit (R.A. Young 196U:65n and T*+f f ) • The category of benefaction appears
to be an important universal and how it is formed could well be added to Wurm's
typological features for Highland languages (l96Ub, et aeq).

53

loHairt Es Set Ti Tense Suffixes

Pr

Pa

NP

Fu

Pf

1 s

-to

- ru

-su

-lua

-eyo

2 s

-te

-ri

-si

-n

-eye

3 s

-ta

-ria

-sa

-lia

-eya

1 dl

- tepa

- ri pa

-5 i pa

-lipa

-epa

2,3 dl

- tepe

- r 1 p i

-sipi

-lipi

-epe

1 pi

-tema

- rl ma

-s 1 ma

-1 Ima

-ema

2,3 pi

- teme

-rlmi

-simi

-1 Imi

-erne

Although it is quite possible to further segment the suffixes in
Chart 5 so that person-number is distinct from tense, or even so that
person is distinct from number, all of these categories are considered
as compound representations and included within the semantic designa-
tion called Tense.

2

Set II tense suffixes regularly combine with stative bases or,

together with vstMP-Rl, with derived stative stems. Some examples are:

(23) rumaa 'to portion out' + 1 sg Pr ->■ rumaa-to

'J am portioning out (for eomeone) '

(24) ada *to look' + 1 sg Pr + vstMP-Rl -^ adaa-to

'J am looking (on behalf of eomeone) *

(25) ria 'to carry (on shoulder)' + 1 pi Pf + vstMP-Rl

-»■ rlaa-ema 'we have carried (on behalf of someone) '

\ox example, consider the following as morphs which mark only tense: -te- (Pr),
-ri- (Pa), -si- (FP), -U- (Fu), and -e- (Pf). Basic person-number forma are:

1

2

3

Sg

-o

-e

-a

Dl

-pa

-pe

-pe

PI

-ma

-me

-me

Morphophonemic rxiles provide surface representations:

-o

-e
-a

+ Pr -»^

-to
-te
-ta

+ Pa -»-

■ru
■ri
■ria

There seems to be little

and so on, where vowel harmony rules convert o -»• u/-r

to be gained by such an exercise: person-number-tense always occur together (or
person-number-tense relationship) and must ultimately be rejoined and specified as
co-occurring obligatorily.

^In a sense, stative bases are inherently altercentric , i.e. the action of the verb
always relates to some implied person or circumstance other than the speaker. Although
benefaction is signsaied by the affixes as a set, stative bases which combine only
with Set II s\if fixes might also simply be called benefactive stems.

54

The first example rumaa Illustrates a statlve verb stem; the other two
examples illustrate statlve stems derived from active verb bases.

Imperative categories are Singular and Non-Singular, Immediate and
Non-Immediate. Imperative suffixes are listed In Chart 6,

Chart 6: Imperative Suffixes

Non-Immediate
Ego Altr

Immediate

Singular
Non-Singular

vBase vBase + vstMp-Rl
-lepaa -tepaa

•pe

Some examples of Set II Imperatives are:

(26) rumaa 'to portion out* + n-imm n-sg -*- rumaa-tepaa

'you all portion it out (for someone) '

(27) ada 'to look' + n-imm sg + vstMP-Rl -► adaa

'look (on behalf of someone) '

(28) ada 'to look' + imm n-sg + vstMP-Rl -»■ adaa-tepaa-pe

'you all look (on behalf of someone else) now'

Set I Tense suffixes are given In Chart 7.

The basic shape of the active verb base to which Set I Tense suf-
fixes are attached can be found in several ways: (1) the shape of the
verb base which results when the regular past tense suffixes occur;
(2) the shape of the verb base which results when the 1 and 2 sg Pr
tense suffixes occur; (3) the shape of the verb base In the environ-
ment of non-immediate sg Set I imperatives.

In other Instances the shape of the verb base changes according to
regular MP rules which will be outlined.

Chart 7: Set I Tense Suffixes

Pr

Pa

RP

Fu

Pf

1 sg

-lo

-wa

-su

-lua

-e

2 sg

1

-e

-e

-SI

-1 i

-e

3 sg

-la

-a

-sa

-Ha

-ea

1 dl

- lepa

-pa

-s t pa

-Hpa

-epa

2,3 dl

-lepe

-pe

-slpi

-1 ipl

-epe

1 pi

- 1 ema

-ma

-sima

- 1 i ma

-ema

2,3 pi

-leme

-me

-simi

-limi

-eme

The obvious gap of *-le is due to a regular sound loss between the East and West
dialects (Cf. Franklin I968),

55

3.2.2.2. Won-Te/tm^na> Suiilxt6

The presence of separate sets of verb suffixes according to the
* medial* or 'final' position of the verb in an utterance is a typolog-
ical feature of New Gulnean Papuan languages. It is also common in
New Guinea Highland languages to treat such clauses in terms of two
intersecting dimensions: the set of verb stems which expound the Pre-
dicate as one dimension, and the syntactic distribution of the Predicate
as the other. In Kewa such aspects of clause relationships are des-
cribed on the Sentence-level (Chapter 6), so that in this section only
the suffix forms which have morphophonemlc rules accompanying them are
outlined. There are many variations of such suffixes, but semantically

they fall Into two main groups : those which mark successive actions

2
and those which mark simultaneous actions.

Actions which are successive may be performed by the same or by dif-
ferent persons. Such actions may also be either altercentric or ego-
centric in benefaction. The suffixes which mark such actions are listed
in Chart 8 (as well as two others which are discussed later). Morpho-
phonemlc rules for the combination of active verb bases and egocentric
successive or simultaneous suffixes (Set I) follow in the next section

Chart 8s Non-Terminal Suffixes

Bene-
faction

Same
Sue.

Person

Sim.

Different' Persons
Sue.

1st

2sg 3 Idl Ipl 2dl

2pl

Ego
Alo

- a
-wa

-ri
-ma

-no

-ina -na -pona -mona -lipina
vstMP-Rl plus above suffixes

- li mi na

3.2.3. Morphophonemic Rules

Rules which specify the shape of the verb base for a particular pat-
tern when that pattern occurs with the Terminal suffixes of Set I are

See especially Wurm (l96Ub:8l) on typological variations. This medial/final cha^
racteristic is widespread in New Guinea. G.B. Milner (1967:735) refers to it as
"Pilhofer*s distinction" in his review of Cowan's GnjvmaA 0^ tht StntanA, Language.
Pilhofer (1933:35) reported "Unselbstadige Formen" or "Satzinnen Formen" (sentence
medial forms) in Kite of the Huon penins\aa. These medial forms had either the same
form and subject throughout ("Durchgangsformen") or changed forms ("Wechselformen")
and different subjects ("Subjektswechsel") .

Chapters k and 6 give examples in Kewa and also references to other New Guinea
language descriptions.

2
See again Pilhofer {ibid) who reported three kinds of actions in Kite: successive
("postemporares"), simultaneous ("kontemporares") and durative ("kontinuatives").

56

given first. The suffix shapes which result are starred if MP suffix
rules are yet to be applied. A general rule of vowel harmony applies
finally to any appropriate tense or verb stem.

The following rule applies to any base of pattern L:

vstMP-R2

XL

X /

Pr
RP
Fu
Pf

^, where Pr j^ l/2sg

Examples are from the base yola *to pull*

(29a) yola (1 sg Pr)
(29b) yo-su (1 sg RP)
(29c) yo-lua* (1 sg Fu)

(1 sg Pf )

(1 dl Pr)

(1 dl RP)

(29d) yo-e*

(29e) yo-lepa

(29f) yo-slpa

(29g) yo-ltpa* (1 dl Fu)

(29h) yo-epa* (1 dl Pf)

'J am pulling '

'X pulled sometime ago'

*I will pull*

'I have pulled'

'We two are pulling'

'We two pulled sometime ago'

'We two will pull '

'We two have pulled'

-1u-"

-to-

-e

-to

-U-

-►

-ta-

-11-

-te-

-e- _

-ta-

To account for the actual shape of the tense suffixes the following
rule must apply:

vafMP-Rl XL + r-1u-"| ("-to-"! (29c)

(29d)
(29e)
(29g)
(29h)

The correct shapes are thus yo-toa (29c), yo-to (29d), yo-tapa (29e),
yo-tepa (29g), yo-tapa (29h). If certain historical facts are taken
into account, which are outside the scope of this grammar, t regularly
replaces 1. This results in ambiguity in the case of the 1 dl Pr and
1 dl Pf suffixes, but only with verb bases of this pattern.
The following rule applies to bases of pattern E or A:

(30a) ri-su (1 sg RP)

(30b) rl-to* (1 sg Pf)

(30c) rl-sima (1 pi RP)

(30d) ri-tema* (1 pi Pf)

(31a) rakepe-su (1 sg RP)

(31b) rakepe-to (1 sg Pf)

(31c) rakepe-sima (1 pi RP)

(31d) rakepe-tema (1 pi Pf)

(32a) ^di-su (1 sg RP)

(32b) ad-e (1 sg Pf)

(32c) adi-slma (1 pi RP)

(32d) ad-ema (1 pi Pf)

'I carried sometime '

'I have carried'

'We all carried sometime'

'We all have carried'

'I husked sometime '

'I have husked'

'We all husked sometime'

'We all have husked'

'I looked sometime '

'I have looked'

'We all looked sometime*

*We all have looked*

57

The variations in the tense suffixes (from those given in Chart 7)
can be accounted for as follows:

(1) by expanding the left-hand side of vafMP-Rl to include pattern E;
this applies to (30b), (30d), (31b), and (31d).

(2) by adding a further vowel harmony rule to provide the correct
surface shapes for (31b and d) :

vafMP-R2

XH

* [:::]

[s::]

The correct shapes are thus ri-tu (30b) and ri-tima (30d).

The combination of active verb bases and the Immediate imperative
suffixes follows the general pattern of vstMP-R2,

XL -► X / Imp

The underlying form of the imperative suffix is -lepaa for all verb
patterns, except that pattern L requires an addition to vafMP-Rl:

XL + -le- ^ -te-

In every case the imperative is marked by -tepaa for altercentric
benefaction (Cf. Chart 6). Examples of each active verb stem morpho-
phonemlc pattern are:

(33) yola *to pull' + n-imm n-sg ->■ yo-tepaa
'You all pull it'

(33a) yola + n-imm n-sg (Altr) -»- yolaa-tepaa
'You all pull it on behalf of someone'

(3^) ria 'to carry' + n-imm n-sg -*■ rla-lepaa
'You all carry it'

(3^a) ria + n-imm n-sg (Altr) -*■ rlaa-tepaa

'You all carry it on behalf of someone'

simply for variants occasionally heard which correspond to
verbs in E. Kewa (I961+:10l^) , e.g. abu-tua 'Twill compeneate'

This rule also accounts
what I called class A-3 verbs

instead of *abu-toa. Other specisa vowel hannony rules are necessary for the verbs
pu 'to go' and fpu 'to come'. Usually these bases conform to Pattern A, with addi-
tional Pattern H rules in some cases. Note however:

"inm iitp ~ (e)po-pe

sg Fu ^ (e)pi-lua
sg Pa (^)p^-a; pu-a
N-Term <^ _(e)p6-no

Alternately, the forms for 'go' and 'come* may be considered pua and epa respectively
(see Table 3). Once this Is done the "irregularities" are resolved by more regular
morphophonemic rules.

(Dpu

58

(35) ta *to hit* + n-imm n-sg -> ta-lepaa
*You all hit it*

(35a) ta + n-imm n-sg (Alo) -»■ taa-tepaa

*You all hit it on behalf of eomeone*

Active verb bases which occur with Non-Terminal suffixes (Chart 8)
also require MP rules. The following rule applies to bases which com-
bine with suffixes marking successive actions involving same persons:

vstMP-R4

l^] ^ [xi°] /-N-Term (suc-sp).

where in the case of examples (40) and (4l) the vowel harmony rule
applies to the vowel preceding the successive-same person suffix, i.e.
-a becomes -u:

(36) pela 'to pull out* + -a -»- pelo-a pfra-wa

'I pulled it out and sat down'

(37) rakepea *to husk ' + -a -»- rakepea-a pfra-wa

'J husked it and eat down*

(38) roa(E) *to pluck' + -a -> ro^-a pfra-wa

'J plucked it and eat down'

(39) pona *to sharpen' + -a -^ pono-a pfra-wa

'I sharpened it and sat down'

(40) rfpfna 'to graep* + - a -»■ rfpfnu-a pfra-wa

*1 graeped it and sat down'

(41) sa**" 'to put* + -a -^ su-a pfra-wa

'I placed it and sat down'

(4la) saa-wa pfra-wa 'I placed it (for someone) and eat
down '

In all of the above examples the tense is signalled by the terminal
suffix -wa (1 sg Pa).

If the persons involved in the successive actions are different the
following rule applies:

vstMP-R5

XL

X

XA

-¥

Xe

■

c XE

■

_

_

„

/ 3rd person

(any number)

This form is reconstructed eu3 *sla 'to put' in pre-Kewa.

59

A variation of vafMP-Rl provides for the only other suffix alternants:

' ' [:ll^!] - [:;:s:] /_2di, pi

Some examples are;

(42) aw^ 'to dig' + -na -*- awc-na pfra-wa

'He dug it and I sat down'

(43) adfa 'to fasten' + -na 4df-na pfra-wa

'He fastened it and I sat down'

(44) yola 'to pull' + -na y6-na pfra-wa

'He pulled it and I sat down*

(45) rodopea 'to break' + -na r6d6pe-na pfra-wa

'He broke it and I eat down'

(46) abula 'to aompensate' + -na abu-na pfra-wa

'He compensated and I sat down'

(47) fpu 'to come' + -na epe-na pfra-wa

'He came and I sat down'

In example (43) the vowel harmony rule must also be applied.

When two successive actions occur the first may be marked for pur-
pose, rather than simply for time (as in previous exaunples). The per-
son, number and tense of the action is again specified in the suffix of
the final verb but the tonal action is a verb phrase (Chapter 5). The
benefactive nature of the purpose suffix may be specified as egocentric
or altercentrlc (note the contrast in (51) and (51a) below).

Some examples are:

(48) ado-la pu-lu 'I am going to see it'

(49) rl-ta pu-a 'He went to carry it'

(50) rikepe-ta pa-llml 'They will go to husk it'

(51) p6go-ta ep-eme 'They have come to jump'

(51a) pogolaa-ta ep-eme 'They have come to jump on behalf of

someone else '

(52) ri!imaa-ta yall-a * In order to ration it out^ he yelled out'

The morphophonemic rule which applies to the combination of active
verb stems and the purpose suffix (basic -la) is a continuation of
earlier rules :

fpu 'to come' -^ cpe / -na. This apparent irregularity can be overcome 83

suggested in note 1 on pe^ge ^7.

60

vstMP-5b

"xe'

"xv~

XL

->•

X

_XA_

_Xo_

/ Pur,

where V Is the penultimate of an XE pattern. If any pattern Is also H,

then o -^ u/H according to regular vowel harmony rules. Additionally,

the purpose suffix variation follows vafMP-Rl such that:

VB-L,E +

la

-ta

(53) pu pfra-wa

(5^) lo pfra-wa

(55) 'ru pfra-wa

(56) su pfra-wa

Note examples (49-51) above.

Gerundive actions which are always by the same person are also part
of a vp and have morphophonemic rules which are the ssime as those of R5,
so that the environment can now be expanded to read: Pur, Ger. Some
examples are :

'Going on^ I eat down*
'Talking^ I eat down'
'Cooking^ I eat down'
'Placing it, I eat down'

(57) rakepe pfra-wa 'Hueking it, I eat down'

(58) abulo pfra-wa 'Compeneating, I eat down'

The gerundive marker can be interpreted either as a zero suffix, or
as the morphophonemic change which takes place, or as both. Example
(53) also illustrates how the gerundive form of the verb to go is often
used to express an on-going or repetitive action.

Verbs denoting altercentric simultaneous actions by the ssime person
or by different persons are marked by the suffixes indicated in Chart 8
and by the accompanying vstMP-Rl:

(59) !aa-ma pfra-wa 'While epeaking on hie behalf, I eat down'

3.2.4. Verb Syntagmemes

Active, stative, or derived stative stems expound the Nucleus of
verb syntagmemes. The Pheriphery is expounded by affixes which are
diagnostic of the type of verb syntagmeme. The Terminal and Non-
Terminal suffixes of the pheriphery are obligatory; other suffixes are
optional within the framework of the particular type of verb syntagmeme,
still other suffixes and clitics are optional to any verb syntagmeme:

NUC

va / + (ASP ) + SET I

dvs

(CAS) +

+ (ASP^) + Set II

61

The Pherlphery can be expounded as follows:

PHERIi

Terni j
N-Term I ,
Term II
\^ N-Term II

y JlRASE + (ASP ) +

/ (CAS) + '

' ABASE '

SBASE

+ ASP

Where If an ABASE is expounded, vst-MP-Rl must be applied. The remaining
functional points for any verb syntagmeme are:

V -^ (NEG) + NUC + PHERI + (ASP^) + ( JmODJ ^*

where CON and MOD denote Sentence Connectors and Sentence Modals , which
are relevant to and discussed in Chapter 6.

The obligatory exponents have already been reviewed. Optional cat-
egories expound NEG, CAS, ASP^, ASP^ , and ASP^. Each of these will now
be dealt with.

3.2.4.1. We^a-tcve

The pre-clitic na- negates the action signalled by certain verb phrases
(see Chapter 5), the complemented action of certain clauses (see Chap-
ter 4), or simply the verbal action of verb syntagmemes. Some examples
are :

(60) na-pa 1 ua

(61) na-ado-la palua

(62) adaal u na-ya-1 I a

{neq-go I will - 'J will not go*)

(neg-see-pur, go I will ^ 'I will not
go to see it')

{longs neg-af firm-he will « 'Be will
not grow tall*)

(63) na-ma-adaalu yaa-Ha (neg-cas-lon^, affirm-he will

will not shorten tt')

'He will

In the latter case, in order for the negative to not attach to a verb,
it must co-occur with the causative pre-clitic. Other examples of the
use of na- follow:

(64) na-pu-lupaa-pe

(65) na-toa

(66) na-ma-adoaa-lia

(67) na-mij-la pu-lu

(68) na-mu-a pua-wa

(69) na-me^-no pua-a

'Don't all of you go now'

'I will not talk'

'He should not oauee (me) to wait.

'I am not going to get it*

'I did not get it and I went'

'I did not get it and he went'

(70) na-m^i-no na-p^a-a 'J did not get it and he did not go'

62

In example (64), na- negates an imperative action; in (65) an action
to be carried out in the future; in (66) also a future action, but one
which will be caused; in (67) and (68) the use of the negative reveals
two different structures and the fact that na- is a pre-clitic rather
than a prefix. In (6?) it negates a verb phrase of purpose while in
(68) it negates only the first action of two successive actions by the
same person. Thus, in order to negate both of two successive actions,
the negator must occur twice, as in (70), or:

(68a) na-mu-a na-pua-wa 'J did not get it and I did not go*

However, there is no counterpart to the verb phrase of (67):

(*67a) na-mu-la na-pu-lu *Not in order to get it^ I am not going'

In other words, na- always moves to the beginning of the constitute
which it negates; in the case of (67) a verb phrase (Chapter 5). Ex-
ample (69) is parallel to (68), but the Identity of the actors change.
The actors are again different in (70) (1 sg and then 3 sg) , but na-
occurs twice and negates both actions.

3.2.4.2. Causative,

As indicated in the formula, if the causative pre-clitic occurs.
Terminal or Non-Terminal suffixes of Set II must occur in the Pheriphery,
The causative also changes a clause syntagmeme to transitive, if it is
already transitive (Cf. Chapter 4).

(71) pfra-pe (stt-imp imm sg =* 'Sit down')

(71a) ma-pfraa-pe (cas-sit-altr imp imm sg = 'Cause

{someone) to sit down )

(72) pfra-wa (ait-l sg Pa == 'I eat down')

(72a) ma-pfraa-ru (cas-e^A-1 sg Pa altr == 'J caused

(someona) to sit down')

(73) ma-rekaa 'Cause it to stand up for someone'

(74) ma-fraa-to 'I am causing it to be cooked on

behalf of someone '

(75) ma-mi naa-saa-tepaa-pe 'You all cause it to be lifted upwards

right now'

3.2.4.3. A^pecX"^

Many of the co-occurrence restrictions of aspects are outlined in
the tagmemic rewrite rules suggested in Chapter 7. Here, the forms of
the various aspect markers and examples are given;

63

(1) -ba(A) '(inceptive)' specifies action that has begun at some
point in time. It is one of the few affixes where morpho-
phonemic rules outlined earlier apply; it belongs to pattern A
and co-occurs with Set I or Set II, Terminal or N-Terminal
suffixes.

(76) fra 'to cook* + -ba + vstMP-R4 + -a (consec sp) «
fra-bo-a... 'having begun to cook it and ...'

(77) Tra-baa-ru-de 'J started to cook it once'

If -ba(A) interrupts a morphophonemic tense alternant which co-occurs
with a base pattern other than XA, the tense is always from those which
co-occur with pattern A:

(78) fra 'to cook' + 1 sg Pf = fr|-tu 'J have cooked it'

(79) fra + -ba + 1 sg Pf = fra-be 'I have begun cooking it*

In other words MP rules apply now to -ba(A), not the verb base.

(2) -pa ^ {oompletive)^ specifies action completed with altercentrlc
benefaction, sometime in the past. That is, it co-occurs only
with past tense suffixes of Set II. Similarly to the * inceptive'
aspect marker, vstMP-Rl applies to this suffix, which optionally
interrupts the vbase and tense suffix:

(80) fra + «pa + 1 sg Pa altr =fra-paa-ru 'I finished cooking

it ifoT someone) '

(3) -la and -ta ^prolongation)^ are forms which mark egocentric
and altercentric benefaction respectively. They occur only
with N-Terminal suffixes which mark different persons:

(81) fra + -la + 1 sg dp = fra-la-no 'I continue cooking it

for sometime and then* , . '

(8la) fra + -ta + 1 sg dp = fra-taa-no 'I continue cooking it for
(altr)

someone for some time and then, , , '

These suffixes also give some idea of simultaneous actions by different
persons, but the first action is prolonged (Cf. also Chapter 6).

(4) -wa Hresidual)^ specifies that some part of the action remains
to be completed. It co-occurs only with Set II Terminal
suffixes:*^

Unless the form which marks permission with N-Terminal suffixes is considered the
same (Cf, Chapter 6). If so, a transformation rule is required to place it following
N-Terminal suffixes.

64

(82) fra + -wa + 1 sg Pa alo ■= fra-waa-ru *I oooked part of it

(for eomeone) '

2
3.2.4.4. ki^ptct

The two aspect markers In this set function as directional aspects
and co-occur only with suffixes of Set II,

(1) -niaa ^{downward motion)^ specified action performed upon some-
thing in a downward fashion:

(83) fra + -niaa + 1 sg Pa altr = fra-nlaa-ru 'I burned it

downward* (aa a hill)

(2) -saa ^{upward motion)^ specified action performed upon something
in a upward fashion:

(84) fra + -saa + 1 sg Pa altr = fra-saa-ru *I burned it upward'

(as a hill)

3.2.4,5* Aipect^

Several aspect markers only follow terminal or N-Termlnal suffixes.
These are:

(1) -de ^ (punatiliar)^ specified action performed at a point in time;
It occurs with a slightly different function with other word
classes and constructions.

(85) fra + 1 sg Pa + -de « fra-wa-de *I oooked it'

(2) -na ^reported seen action)* occurs only with past actions of
Set I suffixes. It is also used for reported speech (Chapter 6).

(86) fra + 3 sg Pa + -na - fra-a-na '3e was aeen to oook it'

(3) -ya ^reported unseen action)* occurs parallel to the form above:

(87) I ra + 3 sg Pa + -ya - Ira-a-ya 'Be is said to have cooked it'

(4) -lo * (desiderative)* expressed a desire that an action takes place.
It mainly follows only N-Termlnal dp suffixes or suffixes which
Indicate purpose (Chapter 5):

(88) fra + Pur + -lo - fra-la-lo '[l] want to cook it'

(89) fra + 3 dp + -lo ■ fri-na-lo 'Be wants to oook it and ... '

(5) -loa * {serialisation)* Indicates that the action Is completed as
one In a series of actions. It follows only N-Termlnal dp suf-
fixes (Cf. Chapter 6).

65

(90) fra + 3 di. . ,^„ - f.i-^^-ioa * After he oooka it, then.,.*

(6) -paa ^ {excluaiDe)^ indicates that the action is exclusive in
nature (Cf. Chapter 6):

(91) fra + 1 pi dp + -pas » fra-mina-paa *We all (alone) should

Qook it*

Because all other clitics that occur with verbs mark a grammatical
function which can better be described on the clause or sentence level,
these are described in Chapters 4, 5 and 6, In the following section
suffixes and clitics occurring with nouns and other word classes are
described.

3.2.5. Noun Syntagmemes

Noun stems consist of bases which are either Simple or Compound.
Compounds are combinations of general nouns which function as a seman-
tic unit and which have the properties of a singular general noun (i.e.
occur with the usual clitics, have the same syntactic settings, per-
turbation patterns with tone, and so on). Compounds often appear to
be derived from other noun phrase patterns. For example, one possible
underlying pattern for compounds is based on N-na N, where -na in a
full phrase type marks the item-as-possessor. Examples such as;

(92) r^pena-unl {tree hone « * stioke*)

(93) y^gaa-frf {chin hair =» 'whiskers ')
(9k) pore-rumu {mountain knee = 'ridge')

(95) pora-uni {road bone « 'trail')

appear to be derived from repena-na unl, yagaa-na frf, and so on.

A further type of noun compound can be recognised often by the fact
that the first noun specifies a generic property for the compound as a
whole:

(96) raf-kutu {axe, bamboo knife « 'bush knife')

(97) aapu-asala {tanket asa la « 'asala laordyline^ leaves ' )

(98) kabe-lapo {pit-pit lapo - 'harden lapo type of cane')

(99) yagi-putf {kunai grass, aggregate = 'grassland')
(100) frf-katl {hair grey « 'elderly')

Two nouns which share the same semantic characteristics and which

can be counted collectively as a unit, can be considered as derived

from a np which contains lapo 'both' as the M exponent. For

num qan

example, compounds such as:

66

(101) ona-ai (woman^ man » 'people')

(102) nogo-naakf igirt^ boy *« 'children')

(103) fnf-agaa {eyea^ mouth « 'face')

(104) kf-komaa (handj upper arm - 'whole arm')

(105) padi-raani (edible pit-pit^ areas = 'vegetables')

(106) mena-irikai ipig^ dog = 'animala')

appear to be derived from ona aa lapo, nogo naakf lapo, and so on.

Body parts are most frequently in a part-whole kind of relationship
and it is Important to note that subordinate semantic relationships
are often due to physiological function (K. Franklin 1963). Thus the
form kfdipaa must be glossed simply 'naila', but the compounds based
upon the form are either aa-kfdipaa (foot nails = 'toenails') or
kf-k'dipaa (hand nails = 'fingernails').

Other compounds which appear similar in form may be derived from
quite different sources;

(107) ona-ada (woman house = 'women's house')

(108) tapa-ada (platform house « 'men's house')

(109) kabe-ada (pit-pit house - 'menstrual hut')

(110) kuku-ada (cook iPidgin Eng,'\ house = 'kitchen')

(111) repena-agaa (fire mouth « 'headlights')

(112) repena-reke (wood atairs « 'ladder')

(113) repena-kaapu (wood dry = 'firewood')

Exan^les (108-110) are derived from a common Modifier-Head functional
phrase pattern, but (107) is more similar to (93-96), based on the
pattern of N-na N. Likewise (111-113)5 although similar in form,
appear to be derived from different patterns.

There are no other characteristics which would serve to distinguish
separate types of nouns. There are also no clitics or suffixes which
occur solely with nouns. There are, however, four which function on
the word- level and which may be considered together because none of
them occurs with verbs.

3.2.5.1. WoKd'Uvzl Clitics

Clitics which primarily attach to word-level tagmemes and which
therefore show no phrase or clause relationships are:

A further qualification is necessary here. Phrases such as ome-nu-mf (to die riaxf
ooll-AGN = 'thoae who are dying') are common, but the full phrase can always be
supplied:

6(ne ona^-nu-mf 'the people who are dying', I consider the use of the collective
clitic -nu in ome-nu-(mO as a contraction of the full phrase-

67

aa- information question' : aa-aa (ques-man « 'what man?')*,
aa-rabu (quea-time - 'what time?'); aa-para (ques-loc =
'where?') \ aa-maapu-nu (ques-^arden-coll = 'what gardens?')

-si ' diminuative quality': yomagae-st {old man-dim =

'a slightly old man'); laapo-si 'two little ones'; adaa-sl
'a slightly big one'*

#, / / '

-nu 'oolleotive' : yomagae-nu 'all of the old men'; ekeraa-nu
{tomorrow -all = 'in the future'); nimu-nu 'all of them ,

-1u * durative quality*: aaraa-lu {father-dwr = 'a family');
pora-lu (rc?ad-dur = 'a long way')

Combinations of word- level clitics which are permissible are:

(11^) mena-sl-nu 'all of the little pigs'
(115) adaa-lu-nu 'all of the long ones'

Because -nu specifies an aggregate, it cannot co-occur with an aj^^
such as: *men^-nu l^^po 'all the pige^ two'.

3.2.6, Other Word Patterns

The three word-level clitics which have been described combine freely
with stems of other non-verb word classes. Each word class will now be
discussed individually.

3.2.6.1 . AdjtctlvalA

In addition to the basic class of adjectives described (Chapter 3)
other adjectives may be derived from verbs. Such forms expound the
Modification function of a np^^^, most often occurring as the relator

of an embedded clause. The derivational clitic is of two basic forms:
egocentric, where the shape of the clitic is determined according to
the underlying morphophonemic pattern of the verb base; and alter-
centric, which is invariably the clitic -e. Examples for each verb
pattern are (Cf. also vstMP-R5):

(a) XL -^ Xne, e.g. yala ^ yane 'the yelling (one)'

(b) Xa H- Xe, e.g. awa -^ awe 'the digging (one)'

(c) XE -► Xni, e.g. ria -> rla 'the carrying (one)'

(d) Xaa -► Xaae, e.g. nfmfnaa -*- nfmfnaae 'the understanding (one)'

In example (c), because XE is also XH (Includes a high vowel), ne -► ni.
Although word-level clitics may attach to the derived adjective, the
preferred pattern attaches the clitic to the noun expounding the Head:

68

(116) rini aa-nu, rather than rihi-nu aa *all of the men who carry*

(117) yane aa-si, rather than yane-si aa Hhe emalleet man who

yelle out*

If the adjective is a cardinal number such as padane *one*, laapo
*two*, repo 'three* and maala *four*y all additional cardinal numbers
are based on multiples of four (Franklin and Franklin 1962, Cf. also
Franklin 1968)- Structurally the forms are numerical noun phrases
and possessive noun phrases, and these are described in Chapter 5.
However, body parts may also be named as ordinal numbers and the base
cardinal numbers may become ordinal. To do this the numerical deriva-
tional clitic -pu * quantifier* is added. It may be added to stems of
certain other word classes as well:

(118) laapo-pu *two of them*

(119) ckita-pu (little finger-quan. = *the first of them*)

(120) adaa-pu (big-qudin, = 'plenty of them')

(121) ade-piS (u?iat-quan. = *how many of them?*)

3.2.6.2. AdvtA.blal^

Derived adverbs have already been mentioned in 3.1.1.^. They
consist of a syntagmeme marked by -rupa. Adverbial clauses and adverbs
which function as clause modifiers are described in Chapter 4.

3.2.6.3. VzldtlcJi

Personal and interrogative pronouns combine with all word-level
clitics except - 1 u :

(122) ne-si * little you*

(123) niaa-nu 'all of ue *

(124) aapf-nu *who all?*

(125) akc-si *th0 little what?*

Interrogative pronouns combine with clitics which function at vari-
ous levels of the grammar. In Chart 9 these are outlined and the
clitic is given a very general gloss. In two instances ake adds the
vowel /a/ before a clitic and in one instance the final vowel of ake
changes in a manner apparently following vstMP-R5b.

69

Chart 9: Interrogatives

Clitics

Animate
5apf 'who'

Inanimate
ake 'what'

-nu

'coll'

aapf-nu

'who all'

ake-nu

'what all'

-ni

'poae'

iipt-ni

'whoee '

-me

'agn'

aapf-mf

'who'

ake-me

'what'

-para

'ben'

aapf -pa ra

'to whom'

ake-para

'about what '

-ne

'adjz '

--

akea-ne

'due to what'

-daa

'obj'

aapf-daa

'due to whom'

--

-lo

'deer'

--

ako-lo

'for what desire
(or purpose) '

Demonstratives, on the other hand, may combine with not only word-
level clitics but also with themselves. Examples with clitics are:

(126) go-sl 'those little ones'

(127) so-nu 'all those up there '

(128) apo-5 i -nu 'all those little ones somewhere over there '

Examples of demonstrative compounds are:

(129) mo-go 'over there (seen) '

(130) s6-go 'up there (seen) '

(131) no-go 'down there (seen)'
(129a) m6-po 'over there (unseen) '
(130a) s6-po 'up there (unseen)'
(131a) no-po 'down there (unseen)'

but not forms such as *mo-so, *mo-no, *go-po. In demonstrative com-
pounds comprising apo, the initial vowel of the stem is lost.

TABLE 3

Verb Paradigms

pu 'to go'

pu 1 u

'I am..,'

pu i

'you are, , , '

pula

'he is.,.'

pul upa

'we two are. . . '

pu 1 up i

'you two are. . . '

pul uma

'we all are, . . '

pu 1 umi

'you all/they are

f pu 'to come '
f pu 1 u

fpui

f pul a
f pu 1 upa
fpui up t
fpui uma
fpui umi

Cont. on next page

70

pu *to

90'

puawa
puae
puaa
puapa

puape

puama
puame

palua
pall
palia
pal 1 pa
palipi
pa 1 i ma
pa n ml

pfsu
pfsi
pf sa
pf s 1 pa
pfslpi
p f s i ma

pfsiml

pe

pe

pea

pepa

pepe

pema
peme

TABLE 3 (Cont. )

' J . . . recently *

'you , . . recently *

*he . . . recently '

'we two . • . recently '

'you two ... recently'

'we all . . . recently '

'you all/they ... recently'

'I will,.. '

'you will. . . '

'he will. . . '

'we two will. . . '

'you two will. . . '

'we all will. . . '

'you all/they will... '

'I ... sometime ago'

'you ... sometime ago*

'he ... sometime ago'

'we two ... sometime ago'

'you two ... sometime ago'

'we all ... sometime ago'

'you all/they ... sometime ago'

'I have. .. '

'you have. . . '

'he has . . . '

'we two have. . . '

'you two have. . . '

'we all have. . . '

'you all/they have...'

f pu 'to come '

epawa
epae
epaa
epapa

epape
epama
epame

epal ua
epal i
epa 1 i a
epa 1 i pa
epa 1 1 pi
epal i ma
epa 1 I mi

fpisu
I p I s 1
I p i sa
f pi s i pa
f p i s i pi
f p 1 s 1 ma
fp i s I ml

epe

epe

epea

epepa

epepe

epema

epeme

3.3.

FURTHER PARAPIGM5

As indicated in the previous section verbs in Kewa occur with two
sets of suffixes, depending on the benefactive nature of the action.
Both sets are now listed with the verb stem ada 'to see' with each of
the five basic tenses.

(1) Present Tense

Ego-centric

1 Sg

ada- 1 o

2 Sg

ada-le"*"

3 Sg

ada- 1 a

1 Dl

ada-lepa

2 Dl

ada- lepe

'/ am looking

'you are

'he is

'we two are. .
'you two are.

Alter-centric

'I am looking for,

'you are

'he is

adaa-tepa 'we two are

adaa-tepe 'you two are

adaa-to
adaa-te
adaa-ta

Cont . on next page

In the West, due to a regular sound change where 1 becomes lost, the form is simply
ada-e. This is Identical, except for tone, with the past tense form: ada-e.

71

Ego-centric (Cont.)
3 Dl ada-lepe 'they two are,**

1 PI ada-lema 'we alt are

2 PI ada-leme ^you all are,,*, '

3 PI ada-leme 'they all are,,,'

Alter-centric (Cont, )

adaa-tepe 'they two are '

adaa-tema 'we all are '

adaa-teme 'you all are '

adaa-teme 'they all are '

(2)

Past Tense

Ego-centric

Alter-centric

E, SK

1 Sg

ada-wa

adaa- ru

-ripu

2 Sg

ada-e

adaa-rl

-ripl

3 Sg

ada-a

adaa- ri a

-rlpa

1 Dl

ada-pa

adaa-rl pa

-pa

2 Dl

ada-pe

adaa-rl pi

-pe

3 Dl

ada-pe

adaa-ripi

-pe

1 PI

ada-ma

adaa-rlma

- r i ma

2 PI

ada-me

adaa-rtml

-riml

3 PI

ada-me

adaa-rlml

-riml

(3) Remote Past Tense

Ego-centric

Alter-centric

1 Sg

ade-su

adaa-su

2 Sg

ade-si

adaa-s 1

3 Sg

ade-sa

adaa-sa

1 Dl

ade-slpa

adaa-slpa

2 Dl

ade-s Ipl

adaa^s 1 p?

3 Dl

ade-s 1 pt

adaa-slpf

1 PI

ade-slma

adaa-s f ma

2 PI

ade-slmi

adaa-s 1 mt

3 PI

ade-s Iml

adaa-slmi

(4) Future Tense

Ego-centric

Alter-centric

1 Sg

ada-lua

adaa-lua

2 Sg

ada-li

adaa-li

3 Sg

ada-1 ia

adaa-Ha

1 Dl

ada-1 ipa

adaa- 1 ipa

2 Dl

ada-llpi

adaa-llpl

3 Dl

ada-1 Ipi

adaa-IIpt

Cont. on next page

"^EK Is identical in Set A» except for 1 Sg» which takes the suffix -kua. Set E Is
also Identical, except for adu-ka (l sg).

72

1 PI

ada-1 ima

2 PI

ada-1 Iml

3 PI

ada-1 (mi

Ego-centric (Cont.) Alter-centric (Cont,)

adaa- 1 ima
adaa- 1 i mi
adaa- I i mi

(5) Perfect Tense

Ego-centric

Alter-centric

1 Sg

ade

adaa-yo

2 Sg

ade

adaa-ye

3 Sg

adea

adaa-ya

1 Dl

adepa

adaa-pa

2 Dl

adepe

adaa-pe

3 Dl

adepe

adaa-pe

1 PI

adema

adaa- ma

2 PI

ademe

adaa- me

3 PI

ademe

adaa-me

^e 2 Sg form in E and SK is adele. The final syllable is lost in WK. following a
regular sound loss. ^

2

This can be derived from a number of factors: by examining the relationship of the
tl^Z^ ir+r''^ ^^ f °^^l°«y of all present day Engan languages which shows that
speakers of the Samberigi language (Sau) must have penetrated southward into the area
before a later movement of Hull speakers to the southwest, but much later than the
Wlru speakers to the southeast.

PACIFIC LINGUISTICS
SzfilzA C - Ho, 53

A KEWA DICTIONARY

WITH SUPPLEMENTARY GRAMMATICAL AND
ANTHROPOLOGICAL MATERIALS

by

Karl J. Franklin and Joice Franklin,
assisted by Yapua Kirapeasi

Department of Linguistics
Research School of Pacific Studies

THE AUSTRALIAN NATIONAL UNIVERSITY